Voltage spike caused by BMS disconnecting

[ianganderton]

2 threads on another forum that seem to be describing a problem where the BMSis disconnecting before the SCC and causing damaging voltage spikes

https://nzmotorhome.co.nz:80/forum/viewtopic.php?t=21237

https://nzmotorhome.co.nz:80/forum/viewtopic.php?t=21330

The problem appears to be the BMS is disconnecting as the voltage goes up and this is causing a spike. I’m not sure why, maybe the solar PV connecting directly through?

The solar charge controller is an epever tracer AN so one that is well regarded. Standard lithium programme settings

The guy thinks it has damaged his diesel heater. The spikes are tripping his inverter which is what originally alerted him to the problem.

 

[Ampster]

Well that is a good argument for not using the BMS as the primary disconnect device. Many posts have suggested that the BMS should be the last resort disconnect. I also have to wonder about the guy with the diesel heater because if the wiring was correct the spike should be a drop in voltage on anything disconnected from the battery. Is it possible that the inverter caps would discharge when power was disconnected? That sounds like an inverter design issue.

 

[ianganderton]

Well that is a good argument for not using the BMS as the primary disconnect device. Many posts have suggested that the BMS should be the last resort disconnect. I also have to wonder about the guy with the diesel heater because if the wiring was correct the spike should be a drop in voltage on anything disconnected from the battery. Is it possible that the inverter caps would discharge when power was disconnected? That sounds like an inverter design issue.

The diesel heater is a 24V model and it’s the voltage spike from the SCC On battery disconnect he thinks has damaged it.

To me it looks like the BMS is set to disconnect earlier than other BMS’s and what epever consider normal.

I’m looking at this battery and SCC combination so it’s particularly interesting. The supplier is local to me.

 

[Ampster]

I’m looking at this battery and SCC combination so it’s particularly interesting.

Keep us informed. Long term it is one of the reasons I am trying to find ways to turn off or turn down the charging current or voltage with my system than having the main contactor drop out the batteries.

 

[HRTKD]

It seems to me that setting the BMS to cut off charge at .01v higher than the cut off for the solar charge controller would alleviate this type of problem. But if the voltage values aren't quite in sync then .01v may not be enough margin for error.

 

[ianganderton]

It seems to me that setting the BMS to cut off charge at .01v higher than the cut off for the solar charge controller would alleviate this type of problem. But if the voltage values aren't quite in sync then .01v may not be enough margin for error.

The BMS in the battery isn’t user changeable

I’m wondering if some of the metering in the BMS is inaccurate and there fore causing the problem

 

[Ampster]

The BMS in the battery isn’t user changeable.
I’m wondering if some of the metering in the BMS is inaccurate and there fore causing the problem

Could be. The solutions are get a new BMS or adjust inverter/charger down and lose some capacity.

 

[ianganderton]

Could be. The solutions are get a new BMS or adjust inverter/charger down and lose some capacity.

These are sold as a drop in and are ‘wrapped up’ in a sealed plastic case so it’s not that easy to change out the BMS

Because it’s sold as a complete unit it is subject to New Zealand consumer law so needs to work as advertised

 

[Ampster]

Ironically it sounds like the BMS and battery are working as advertised. It is the surge from the SCC that caused the problem with the diesal heater.

 

[HRTKD]

The BMS in the battery isn’t user changeable

I’m wondering if some of the metering in the BMS is inaccurate and there fore causing the problem

Change the solar charge controller's threshold to be lower than the BMS.

 

[HaldorEE]

Since the problem is in the Diesel Heater, I would try adding a TVS and a small cap across the DC input to the heater. A series choke may end up being required also.

I can't open the OPs links, what voltage is the battery?

That's the problem with buying cheap equipment, they leave out all the "unnecessary" components that higher priced equipment include. I am going be buying the same cheap heater for my van, so good to know the DC input needs some extra protection.

 

[HaldorEE]

Something like this plus a TVS with the right voltage rating.

Universal Car Radio Audio Power Wire Engine Motor Noise Filter Suppressor 12V | eBay

Find many great new & used options and get the best deals for Universal Car Radio Audio Power Wire Engine Motor Noise Filter Suppressor 12V at the best online prices at eBay! Free shipping for many products!

www.ebay.com

That one would work on a 12V system, but the cap voltage rating is too low to work on a 24V system. A 50V cap would have cost pennies more. Cheap bastards.

I might buy it anyway and replace the cap myself.

 

[HaldorEE]

Love the description of that noise filter. That is precisely what I want to do. Eliminate the car inside the car and keep my water ripple from being interfered with.

Item description
"12V Car Vehicle Stereo Radio Audio Power Wire Engine Noise Filter Suppressor. Eliminate the car inside the car on the car audio power interference caused by noise, but also can effectively inhibit the computer display water ripple interference!"

I used to work with a very smart Chinese engineer. His favorite saying was "There is a thing there!" Course his English was 1000000 times better than my Chinese.

 

[gnubie]

If you've ever had another car enter your car you know just how noisy its engine can be so it's always a good thing to remove the noise from the other car.
?

 

[HaldorEE]

If you've ever had another car enter your car you know just how noisy its engine can be so it's always a good thing to remove the noise from the other car.
?

And I bet it will ripple your water too.

 

[ianganderton]

Some additional information about the Lithium Valley batteries

Quote: “ I have two of these Lithium Valley Batteries so followed those threads with interest. Even though the spec sheet indicated the battery BMS cut off voltage is 14.6 volts ,in reality it is only 14.3 volts. I have confirm this with my own batteries. The issue was with a solar charger which was set up to charge to 14.6 volts. The BMS would kick on at 14.3 volts essentially causing an battery open circuit. This caused the solar controller to also go into a no load situation which resulted in voltages from the controller to raise above 17volts. Keeping charge voltages below 14.2 volts has solved the issues.”

How does this affect battery performance? Is there much storage lost?

14.6V / 4 = 3.65v per cell
14.3V / 4 = 3.575V per cell
14.2V / 4 = 3.55V per cell

I think the retailer of the battery should be doing something about a significant difference in performance of the BMS. It’s causing significant problems with SCCs using the Lifepo4 charge settings they have recommended on units (epever) they have sold with them

 

[gnubie]

There's another possibility, the accuracy of the charger itself. I can't speak for the particular charger in use but I regularly see cheaper chargers with errors up to +/- 0.2V with the cheapie end of the market being worse maybe +/- 0.4V. Even for chargers that offer a calibration offset there can be shortcomings in the design that mean the error is still there, just reduced.

It's entirely possible that the BMS is working correctly and the problem actually is the charger. Checking the actual voltage the charger is targetting in absorption mode will tell, ie set the absorption charge to below where the BMS trips and measure the voltage with a decent multimeter. A cheapie meter probably has unimpressive calibration itself. It must be done when there is little current flowing to avoid voltage drop. Check the voltage at the battery's terminals.

While charging, ie the charger is pulling the battery up over its resting voltage, but with minimal current flowing to keep it at 3.55V, it is somewhere between 80-90% capacity. It's temperature and specific chemistry dependent. Unless you know the exact chemistry of your lifepo4 cell this is probably as close as you can get using cell voltage alone.

 

[jensenq]

Since the problem is in the Diesel Heater, I would try adding a TVS and a small cap across the DC input to the heater. A series choke may end up being required also.

I can't open the OPs links, what voltage is the battery?

That's the problem with buying cheap equipment, they leave out all the "unnecessary" components that higher priced equipment include. I am going be buying the same cheap heater for my van, so good to know the DC input needs some extra protection.

I like @HaldorEE 's idea of adding a clamp circuit to deal with spikes coming from the solar charge controller if the BMS disconnects. I have seen this with both PWM and MPPT solar charge controllers, causing downstream overvoltage faults at a sine wave inverter and stressed or fried LED house lights, during set-up and configuration of the max charge voltage of the solar charger, and the BMS disconnect parameters. I get the impression that the designers of these charge controllers neglect the rather recent development of Lithium+BMS battery banks suddenly disappearing as a charging load. The legacy lead acid banks never do that, and properly configured Lithium+BMS banks shouldn't either, but since it can happen, it eventually will happen. You might have peace of mind that your Lithium bank won't ever catch fire, but when your BMS disconnects to prevent that, you don't want downstream loads to fault or fry.

What I'd like to know is what the regulation recovery time is for this event for a given MPPT controller -- i.e. how long the voltage spike can last? It would be nice to study these events on a storage scope in a test environment without any expensive loads downstream that could fry. Then you could design a proper transient suppression circuit to deal with it, like @HaldorEE has begun to do on this thread. Oh, and the SCC designers should make sure their outputs can never exceed a peak voltage for their configured system voltage. And the SCC's that automatically detect the system voltage might not be trustworthy because of BMS disconnect eventualities.

Maybe this would be a fun test for @Will Prowse to reproduce in his shop. Has he seen inverters fault or other loads complain or fry when the BMS disconnects the battery during full solar charge? This is a very different test case than trying to fry an SCC by disconnecting the battery suddenly.

 

[sajohnson]

Some additional information about the Lithium Valley batteries

Quote: “ I have two of these Lithium Valley Batteries so followed those threads with interest. Even though the spec sheet indicated the battery BMS cut off voltage is 14.6 volts ,in reality it is only 14.3 volts. I have confirm this with my own batteries. The issue was with a solar charger which was set up to charge to 14.6 volts. The BMS would kick on at 14.3 volts essentially causing an battery open circuit. This caused the solar controller to also go into a no load situation which resulted in voltages from the controller to raise above 17volts. Keeping charge voltages below 14.2 volts has solved the issues.”

How does this affect battery performance? Is there much storage lost?

14.6V / 4 = 3.65v per cell
14.3V / 4 = 3.575V per cell
14.2V / 4 = 3.55V per cell

I think the retailer of the battery should be doing something about a significant difference in performance of the BMS. It’s causing significant problems with SCCs using the Lifepo4 charge settings they have recommended on units (epever) they have sold with them

I've been experiencing this exact problem. Equipment used:

* Magnum Energy MS2000 inverter/charger
* Magnum ME-ARC remote
* 2 each Lion UT1300 LFP batteries
* Victron SmartShunt

The MS2000 worked flawlessly with the Crown GC2 batteries we had previously, but it does not play well with the Lion LFP batteries -- at least not when set per Lion and Magnum's instructions. I had the charger set for 100% (~100A) -- 50A per battery -- which is what Lion recommends. The charge current was just over 100A (per the ME-ARC). The voltage was 14.1-14.2. The SmartShunt read just 77A and 14.05V. I haven't figured out why there is such a large difference in the current readings.

More significant is that although the batteries' SOC was just 40%, they only accepted a charge briefly and then the BMS opened up and stopped current flow. There was a voltage spike -- the red LED on ME-ARC flashed, and there was "High Battery" fault code: H1 D- 17.2V -- 0A. I could also see the 17.2 volt spike on the SmartShunt app on my phone. That happened a couple more times and then I shut the charger down.

I spoke with a tech at Sensata/Magnum and he had me enter the following settings: 30A AC in / 60Vac dropout (was 80) / 'Custom' charger setting / Absorp done = 100A / Max charge 90% (was 100%) / Max charge time = 1 hour / Final charge stage 'silent' / EQ reminder 'Off' / Absorb = 14.6V / Float = 13.6V (was 14.6V per Lion's instructions) / EQ = 14.6V / EQ done time = 0.1 hours.

He told me to turn the charger on -- when I did, the same thing happened. The batteries took a charge for a while and then the BMS opened up and I got the voltage spike and another fault code. As I recall when it began to charge again it charged for a couple minutes before shutting down.

Also, when charging stops, the 120Vac pass-through in the MS2000 was usually interrupted for maybe 15 seconds. There should of course be 120V at the output at all times. I'm not sure what caused the ATS to open up, unless it was the 'high battery' fault.

I plan to attempt to balance the cells by setting absorb and EQ to 14.0V and letting the batteries charge that way for a few hours (if the BMS will allow it). Then I'll try 14.2V for absorb and EQ for ordinary charge cycles.

Any thoughts and/or theories welcome.